Over the last two decades, we have seen a majority of the most destructive 20 fires in the history of California. Wildland fires are becoming more frequent and fire behavior is changing to be more destructive due to a combination of natural and anthropogenic factors, which makes applied fire research very important and time critical. The WIFIRE Commons enables AI-driven societal and scientific wildland fire applications through data and model sharing. Our interdisciplinary team of fire science, data, AI, cyberinfrastructure and science to practice experts will build upon the successes and existing partnerships formed as part of the WIFIRE project previously funded by the NSF. These partnerships span a growing number of academic institutions, public agencies and utilities, as well as cross-sector industry partners.

Proactive fire management to reduce destruction by wildfire can be optimized through novel integration of data and models, but it has not been achieved yet due to inconsistencies between data, modeling needs, and the actual sensitivities of fire behavior in various conditions. Science-based decisions for fire mitigation, preparedness, response, and recovery need improved dynamic data-driven models. As a solution to this problem, WIFIRE Commons will scale our earlier work with continuous community-driven capability enhancement, data curation and access. Our main objective is to build and enable artificial intelligence innovations that power wildland fire science and its proactive application to operational use for mitigation, planning, response, and recovery.

The WIFIRE Commons framework has the potential to advance wildland fire innovations not just through data and model sharing, but also through the use of AI to continually refine and optimize the use of available wildland fire data and models to protect lives and property at the wildland urban interface (WUI), as well as natural resources. We will enable development of novel AI methods that integrate data and modeling with an evolving suite of tools in a flexible multi-fidelity application translation framework. This Commons infrastructure will catalog, curate and integrate data and models for AI-driven fire science, maintain open programmatic access to data in a cloud-compatible form that can be integrated into the AI process through a gateway interface, and ensure provenance of data and models over time. Through this AI-enabled smart data/model integration, our long-term aim is to transform the agility of science-based wildland fire decision making, allowing for new kinds of models and data to be assimilated rapidly and allowing users to understand levels of uncertainty originating from models or data.